阵列数据库系统的存储块分割策略研究

The research of chunk segmentation strategy for array database system

阵列数据库系统是存储和分析大规模科学数据的常用技术方案。目前主流的阵列数据库中存储块分割策略采用固定边长作为块的边界,若边长过大会增加查询分析时定位Cell的时间,反之则产生过多的小块增加内存开销。本文提出一种改进的Chunk边长分割算法CLD,其通过减少读取数据时的磁道数以及预取技术提高阵列数据库系统的性能。在阵列数据库系统Sci DB集群上的实验表明,在最优情况下系统性能提升了10.9%。

Array database is often used for massive scientific data storage and analysis. The major array database systems primarily employ the fixed length strategy for chunk segmentation. If the length of the chunk is too big, it will increase the overhead of cell location during query analysis. On the contrary, it will make too many small chunks and increase the memory overhead. This paper proposed a chunk length divide（CLD） algorithm to improve the performance of array database by reducing the number of track and prefetching technique when the data is read. Based on the evaluation of array database SciDB cluster, our approach lead the performance of SciDB improved about 10.9% under the optimal conditions.